• Non ci sono risultati.

The end of the Lago-Mare time in the SE Valdelsa Basin (Central Italy): interference between local tectonism and regional sea-level rise

N/A
N/A
Protected

Academic year: 2021

Condividi "The end of the Lago-Mare time in the SE Valdelsa Basin (Central Italy): interference between local tectonism and regional sea-level rise"

Copied!
29
0
0

Testo completo

(1)

Abbazzi L., Benvenuti M., Ceci M. E., Esu D., Faranda C., Rook L. & Tangocci F. 2008. — The end of the Lago-Mare time in the SE Valdelsa Basin (Central Italy): interference between local tectonism and regional sea-level rise. Geodiversitas 30 (3) : 611-639.

ABSTRACT

A multi-disciplinary study has been carried out on a Late Neogene succession exposed in the SE portion of the Valdelsa Basin (Strolla Creek, Central Italy). Th e succession consists of upper Tortonian-Messinian sediments, unconform-ably overlain by uppermost Messinian-Pliocene deposits, accumulated in al-luvial, lacustrine and marine environments. Previous studies on this succession hypothesized a sudden marine fl ooding of the upper Messinian Lago-Mare realm in the earliest Zanclean in apparent agreement with the Mio-Pliocene transition

Laura ABBAZZI

Dipartimento di Scienze della Terra, Università di Firenze, via G. La Pira 4, I-50121 Firenze (Italy)

Marco BENVENUTI

Dipartimento di Scienze della Terra, Università di Firenze, and Istituto di Geoscienze e Georisorse, C.N.R, c/o Dipartimento di Scienze della Terra, Università di Firenze,

via G. La Pira 4, I-50121 Firenze (Italy)

Maria Elena CECI

Dipartimento di Scienze Geologiche, Università di Roma Tre, Largo San Leonardo Murialdo 1, I-00146 Roma (Italy)

Daniela ESU

Dipartimento di Scienze della Terra, Università di Roma “La Sapienza”, Piazzale Aldo Moro 5, I-00185 Roma (Italy)

Costanza FARANDA

Dipartimento di Scienze Geologiche, Università di Roma Tre, Largo San Leonardo Murialdo 1, I-00146 Roma (Italy)

Lorenzo ROOK Francesca TANGOCCI

Dipartimento di Scienze della Terra, Università di Firenze, via G. La Pira 4, I-50121 Firenze (Italy)

interference between local tectonism

(2)

MOTS CLÉS Messinien supérieur, Zancléen inférieur, Lago-Mare, associations fossiles non-marines, analyses de faciès, Bassin de Valdelsa, Italie centrale.

recorded elsewhere in the Mediterrranean Basin. Data from this study suggest a more complex stratigraphy recording a local depositional evolution possibly driven by interference between uplift, relief denudation and eustacy. Th e sedi-ments encompassing the Mio-Pliocene transition have been included into the Borro Strolla synthem, furtherly subdivided into three sub-synthems. Th e Borro Strolla synthem has been dated as uppermost Messinian-earliest Zanclean based on the integration of physical stratigraphy, facies analysis and biochronology of non-marine fossil assemblages in part documented for the fi rst time. Th e oc-currence in the lower Borro Strolla Synthem of the gerbil Debruijnimys sp. and the murid Stephanomys aff . S. donnezani (Depéret, 1890) with Pliocene affi nity represents the most important novelty in the composition of a mammal fauna otherwise similar to other Italian late Messinian sites. Th e co-occurrence of ter-restrial molluscs from the same sediments gives further biochronologic constrains supporting the calibration of the Borro Strolla synthem to the latest Messinian-earliest Zanclean. Th e Borro Strolla stratigraphic section then is discussed in the framework of local to regional events which marked the transition from the Messinian to Pliocene.

RÉSUMÉ

La fi n du Lago-Mare dans la partie SE du bassin de Valdelsa (Italie centrale) entre tectonique locale et remontée du niveau marin régional.

Une étude multidisciplinaire a été réalisée sur les dépôts néogènes (alluviaux, lacustres et marins) du sud-est du bassin de Valdelsa (Italie centrale). Ces sédiments, datés du Tortonien supérieur-Messinien inférieur, sont surmontés, en discontinuité, par des sédiments du Messinien supérieur-Pliocène. Cette étude suggère que ces dépôts résultent de l’interaction de phénomènes locaux tels que la surrection et la dénu-dation des reliefs ainsi que l’eustasie plutôt que du changement drastique dû à la transgression du Zancléen inférieur sur le Lago-Mare, enregistrée en Méditerranée. Ces sédiments, déposés à la transition du Mio- Pliocène correspondent au synthème de Borro Strolla (qui comprend trois sub-synthèmes). Il a été daté Messinien supé-rieur-Zancléen inférieur grâce à l’analyses de coupes, d’études sédimentologiques (étude de faciès), et de la biochronologie établie sur les associations de faunes non-marines, montrant elles-mêmes de nouveaux éléments. La découverte, dans la partie inférieure du synthème de Borro Strolla, du gerbillidé Debruijnimys sp. et du muridé Stephanomys aff . S. donnezani (Depéret, 1890) à affi nité pliocène est un élément nouveau dans cette association faunistique qui correspond par ailleurs à celle observée en Italie dans les sites du Messinien supérieur. D’autre part, la pré-sence de mollusques terrestres dans ces mêmes sédiments permet aussi d’attribuer le synthème de Borro Strolla au Messinien supérieur-Zancléen inférieur. Les attri-butions stratigraphiques établies pour le synthème de Borro Strolla sont corrélées aux événements locaux et régionaux enregistrés à la limite Messinien-Pliocène.

INTRODUCTION

Th e period encompassing the latest Miocene- earliest Pliocene was marked in the Mediterranean region

by a well-known sequence of dramatic palaeo-environmental changes though not unanimously interpreted in terms of sedimentary successions, including thick evaporites, chronology and forcing

KEY WORDS Late Messinian, early Zanclean, Lago-Mare, non-marine fossil assemblages, facies analysis, Valdelsa Basin, Central Italy.

(3)

factors (see recent review in Rouchy & Caruso 2006). A late Messinian catastrophic sea-level fall aff ecting the whole Mediterranean Basin is generally accepted following the seminal papers on the Messinian Salin-ity Crisis (MSC) (Hsü et al. 1973, 1975, 1978; Cita

et al. 1978). Nevertheless, in the last 15 years such

a dramatic sea-level fall has been considered to have occurred either diachronically or synchronously in marginal and deeper areas, in a time-span ranging from about 7 to 5.6 Ma (see review of alternative time and facies models for the MSC in Rouchy & Caruso 2006 and Manzi et al. 2007). At the end of the Messinian, between 5.59 and 5.33 Ma (Krijgsman

et al. 1999), a dilution of a hyperaline Mediterranean

Basin determined the formation of a low-salinity sea known as the “Lago-Mare” realm (Ruggieri 1962, 1967a, b). Th e sedimentary successions referred to this latest Messinian setting, widely recognized in many circum-mediterranean basins and including resedimented and primary evaporites and clastic deposits, bear a typical palaeontological signature characterized by molluscs, ostracods and dynocysts with a marked Paratethyan affi nity (for a review see Orszag-Sperber 2006). Th is feature has fed a debate which is still in large part open, around the latest Messinian connections among the Paratethys, the Mediterranean and the Atlantic basins (Orszag-Sperber 2006). Finally, the restored connection between the Atlantic Ocean and the Mediterranean Basin, at the beginning of the Pliocene, determined the sudden replacement of the Lago-Mare realm with open and deep marine conditions. Th ese events had a great impact on the structure and composition of the circum-mediterranean biotas forcing large-scale fau-nal exchanges, local extinction, palaeobio geographic isolation, etc. (Benson 1976, 1984).

Such a sequence of events was mostly controlled by tectonically-driven sea-level changes, related to the temporary closure and re-opening of the Gibraltar Strait, and by high-frequency climatic fl uctuations (Bertini 1994, 2006; Butler et al. 1995; Orszag-Sperber et al. 2000; Rouchy et al. 2001; Griffi n 2002). Th e latter determined cyclo-themic development of the Messinian successions exposed in the Mediterranean area. Th e active tec-tonic setting related to the late collision between the Africa and Europe plates, was also aff ecting

at local scales the circumediterranean margins with consequent control on the depositional pat-terns in several Messinian basins (Boccaletti et al. 1990b). Th e basins developed on the western side of the Northern Apennines (Central Italy) bear the sedimentary record of these events outlining the interference between local tectonism and sea-level changes occurred in the late Messinian. Angular unconformities within the evaporite-bearing suc-cessions and resedimented evaporites, in fact, record the active uplift and erosion of the basin margins across and following to the Salinity Crisis (Testa & Lugli 2000). Th e Mio-Pliocene transition has been described in many Tuscan basins as stratigraphi-cally concordant (Bossio et al. 1993) equating the local Lago-Mare replacement by normal marine condition to the sudden, catastrophic, refi lling of the Mediterranean Basin. Nevertheless, locally this transition may be stratigraphically more complex hinting to interplay of eustacy and tectonism on the depositional dynamics.

Th is paper summarizes a case study in which a previously reported concordant contact between Lago-Mare and open marine sediments is only apparent. Specifi cally, data from integrated strati-graphic and palaeontological analyses carried out in the SE Valdelsa Basin (Strolla Creek area, central Tuscany), indicate complex stratigraphic relations among marine and non-marine deposits which escape the picture of the Mio-Pliocene transition recorded elsewhere in the Mediterranean area.

GEOLOGICAL SETTING OF THE VALDELSA BASIN

Th e study area is located in the southeastern end of the Valdelsa Basin (Fig. 1), a 25 km wide and 60 km long NW-SE trending depression, bounded by the Monte Pisano-Poggio del Comune ridge to the SW, and the Monte Albano-Monti del Chianti ridge to the NE and SE and fi lled with more than 2000 m thick Neogene and Quaternary sediments (Ghelardoni et al. 1968).

Th e substratum of these deposits (Fig. 1B) is com-posed on the SW of Late Paleozoic-Jurassic phyllites, quartzites, evaporites and carbonates (Adria basement

(4)

and Tuscan units) thrusted by sandstones, limestones and claystones, Late Cretaceous- Paleogene in age (Ligurid units). Oligocene Macigno sandstones (Tuscan units) dominate on the NE side, whereas limestones and claystones of Ligurid units form the eastern and southeastern margins of the basin.

Th e tectonic origin of the Valdelsa Basin, as well as of the several basins west of the Northern Apennines, feeds a debate between supporters of “extensional” (Trevisan 1952; Sestini 1970; Martini & Sagri 1993; Carmignani et al. 1994; Elter & Sandrelli 1994) and “compressional” (Bernini et al. 1990; Boccaletti et al. 1990a, 1995; Boccaletti & Sani 1998; Bonini & Sani 2002) models. Th ese basins (Fig. 1A), fi lled with up-per Miocene-Pliocene fl uvio-lacustrine and shallow-marine deposits (central basins sensu Martini & Sagri 1993) or exclusively with middle Pliocene-Pleistocene fl uvio-lacustrine and fl uvial deposits (peripheral basins

sensu Martini & Sagri 1993) a few hundreds meters up

to 2500 m thick, are, in fact, alternatively interpreted as grabens/half-grabens or thrust-top basins.

Th e basin-fi ll development during the Messinian was regulated by a major regional relief, the Mid Tuscan Ridge (Fig. 1A), representing a signifi cant physiographic divide. Several basins developed on the east of this ridge, such as the Valdelsa Basin, did not experience evaporitic deposition during the Salinity Crisis maintaining fully terrestrial environments up to the early Pliocene. On the contrary, west of this ridge, marine ingression since early Messinian favoured accumulation of gypsum and salts.

From a stratigraphic point of view the Neogene sediments exposed in the Valdelsa Basin, mostly of Pliocene age, have been recently (Benvenuti & Degli Innocenti 2001) included in a stratigraphic frame-work based on fi ve main unconformity-bounded units (synthems: International Subcommission on Stratigraphic Classifi cation 1994).

In the Strolla Creek valley a succession including the Miocene-Pliocene transition has been described by Bossio et al. (1993, 2001; Fig. 2). Th e basal term of this succession consists of non-marine clays (“unit A” or “Casino clays” respectively by Bossio et al. 1993 and 2001) ascribed to the uppermost Messinian on the base of the non-marine ostracod and mollusc as-semblages. According to Bossio et al. (1993, 2001) this unit is conformably overlain by marine clays (“unit B”

or “Argille Azzurre-lower part”) referred to the lower-most Pliocene Discoaster variabilis-Sphaeroidinellopsis biozones. Th is unit is in turn unconformably capped by paralic gravels passing upward to shallow marine sands (“units C-D” or “Conglomerati di Casa Stieri”) and clays (“unit E” or “Argille Azzurre-upper part”) referred to the middle Pliocene. Th e local succes-sion is topped by shallow marine sands (“unit F” or “Sabbie di Talciona”), in erosive contact with “unit E”, ascribed to the middle Pliocene as well on the basis of the fossiliferous content.

According to Benvenuti & Degli Innocenti (2001) the middle Pliocene units C-F correspond to the Certaldo (“units C-E”) and Ponte a Elsa (“unit F”) synthems whereas unit B may represent part of an older synthem (Borro Strolla synthem, see below) within the Valdelsa Basin Pliocene succession.

LOCATION AND STRATIGRAPHY OF THE STROLLA CREEK AREA SE VALDELSA BASIN

Th e study area is located near Poggibonsi, 45 km SE to Florence (Fig. 3). Th is is a hilly area stretching along the western side of the Chianti Mounts, drained by two small tributaries of the Elsa river, the Carfi ni and Strolla creeks respectively. Along the Strolla Creek, good exposures of the Neogene deposits are available thanks to extensive quarrying of gravel and sand.

On the NNE of the study area the Mio-Pliocene deposits rest unconformably on the pre-Neogene substratum forming the Chianti Ridge. Upper Mio cene sediments, unconformably overlain by the Pliocene succession, crop out to a limited extent in the study area but they are widely exposed in the nearby Casino Basin (Lazzarotto & Sandrelli 1977; Bossio et al. 2001). Th e fi eld recognition of unconformities or inferences on unconformable transitions supported by facies relations and bio-chronologic considerations allowed to subdivide uppermost Messinian-earliest Zanclean deposits into three unconformity-bounded units consisting of various lithofacies associations (see Benvenuti & Degli Innocenti 2001 for a similar approach) in-dicative of diff erent paleoenvironments developed in this area.

(5)

Florence Siena Leghorn Grosseto 50 km Lucca G C M FPP Late Burdigalian-Tortonian deposits Magmatic rocks Late Tortonian-Messinian deposits Pliocene deposits Early Pleistocene deposits Middle Pleistocene to recent deposits Pre-neogene units Arezzo Arno River Ser chio River Alluvial deposits (Pleistocene-Olocene) Marine and continental deposits (Plio-Pleistocene) Marine and paralic deposits (lower-middle Pliocene) Fluvio-lacustrinel deposits (upper Miocene)

Late Paleozoic-Jurassic basement Ligurid limestones

Ligurid and Subligurid units

Tuscan units Tuscan

Metamorphic units Substratum

Axis of anticline Isobates of the top of the substratum (m) A B Bolsena Lake N VP TF VE SR VC BC Alb Cec Ser Vol Vel VA Mbam 20 km CHIANTI RIDGE Rad PRA TOMAGNO ITALY 200 0 2000 1000 1000 2000 500 50 0 500 500 Casino Ba sin Elsa River Pesa River Val d'Elsa Basin Iano - Poggio de l Comune Ridge PISANI MOUNTS study area Cas Rib N N CHIANTI MOUNTS

MID TUSCAN RIDGE

FIG. 1. — A, Schematic distribution of Neogene-Quaternary basins west of the Northern Apennines (dotted line borders area in B); B, schematic geology of a portion of central Tuscany including the Valdelsa and the Casino basins (after Benvenuti & Degli Inno-centi 2001). Isobates of the top substratum after Ghelaroni et al. (1968). Abbreviations: Plio-Quaternary basins: C, Casentino Basin; FPP, Firenze-Prato-Pistoia Basin; G, Garfagnana Basin; M, Mugello Basin; VA, Upper Valdarno Basin; VC, Valdichiana Basin; Mio-Quaternary basins: Alb, Albegna Basin; BC, Baccinello-Cinigiano Basin; Cec, Cecina Basin; Cas, Casino Basino; Mbam, Monte-bamboli Basin; Rad, Radicondoli Basin; Rib, Ribolla Basin; Ser, Terrazzano Basin; SR, Siena-Radicondoli Basin; TF, Tora-Fine Basin; VE, Valdelsa Basin; Vel, Velona Basin; Vol, Volterra Basin; VP, Viareggio-Pisa Basin.

(6)

UM2 UM1 T ortonian Piacenzian Zanclean Middle Lower PLIOCENE

Bossio et al. 2001 This study

B Pst BS3 BS2 BS1 Pa Pa PS McL1 MsQ MbG MaC Mc McL MaF Mc Mc D E F C Bossio et al. 1993 MIOCENE Upper Messinian A CR PE

FIG. 2. — Comparison among different stratigraphic framework established for the Neogene deposits in the Strolla Creek area. Legend for codes: A-F, see text. Abbreviations: BS1-3, sub-synthems within the Strolla Creek synthem; CR, Certaldo synthem; MaC, Ca-sino Clays; MaF, Torrente Fosci Clays; MbG, Grotti Breccias; Mc, Caprareccia Sandstones; McL, Podere Luppiano Conglomerates; McL1, Lilliano Conglomerates; MsQ, Quercegrossa Sandstones; Pa, Argille Azzurre; PE, Ponte a Elsa synthem; PS, Casa Stieri Con-glomerates; Pst, Talciona Sand; UM1, upper Tortonian-lower Messinian lacustrine deposits; UM2, Messinian Lago-Mare deposits.

UPPER MIOCENEDEPOSITS

(UPPER TORTONIAN-MESSINIAN)

Th e upper Miocene deposits are visible in small outcrops along and around the Strolla Creek (Fig. 3) and are represented by muddy sediments includ-ed within two units (UM1 and UM2). Despite stratigraphic transition is not directly visible, an unconformable contact between the two units is here suggested (see below):

– UM1: light brown-grey mostly massive silty clays bearing calcareous nodules, vegetal debris, ostracod valves and scanty mollusc remains. Th ese deposits, observed exclusively in limited outcrops NE and SE of the Strolla Creek, are correlatable with those

ascribed by various authors to the fi rst lacustrine event in the Casino Basin (“lower lacustrine cycle”, Lazzarotto & Sandrelli 1977; “Argille del Torrente Fosci”, Bossio et al. 2001). Ostracod assemblages (see below), in fact, point to a late Tortonian-early Messinian age, in agreement with the biostrati-graphic calibration provided by these authors; – UM2: greenish-grey clays and silty clays with abundant organic matter, vegetal debris, calcareous nodules and rich oligohaline mollusc and ostracod assemblages (see below). Th ese muddy deposits, in the outcrop 1.5 m thick, are prevalently massive although thin lamination has been locally observed. Th e mollusc fauna is either sparse in the sediment or

(7)

FIG. 3. — Schematic geological maps and cross-section of the Strolla Creek area. Abbreviations: g, gravelly lithofacies; m, muddy lithofacies; s, sandy lithofacies.

Str olla Cr eek Piaggiole A quarry Piaggiole B quarry 270 m A A’ 190 m BS3 BS3 BS1 BS2 BS3 BS3 BS3 250 225 200 200 200 225 175 17 17 15 15 7 15 15 16 16 275 BS1 BS2 BS3 BS3 BS3 A’ BS3 BS3 A BS3 BS3 200   m UM2 Certaldo synthem (uppermost Zanclean-lowermost Piacenzian) Certaldo synthem Ponte a Elsa synthem Alluvial deposits normal fault active quarry inactive quarry UM2 deposits Borro Strolla synthem Borr o Str olla Borr o Str olla BS1 BS2 BS3 m g m s g

Ponte a Elsa synthem

UM2 deposits (Messinian)

Ligurid substratum

sample for ostracod analysis fault

UM1 deposits (uppermost Tortonian-lowermost Messinian) Borro Strolla synthem (uppermost Messinian-lowermost Zanclean) Recent   alluvial   deposits

Bozzagone Fizzano Villa rosa Vegi Carfin Creek Poggio di Villore San Fabiano 3 1 2 300 350 250 250 200 200 250 Strolla Cre ek Piaggiole B quarry Piaggiole A quarry

(8)

FIG. 4. — Panoramic view of the sub-synthem BS3 showing the cyclic stacking pattern of fl uvial gravely-sandy lithofacies associations. The outcrop, the Sefi Quarry face affected by a normal fault, is about 45 m high.

talus BS3-sandy lithofacies talus BS3-sandy lithofacies BS3-gravelly lithofacies BS3-gravelly lithofacies BS3-sandy lithofacies BS3-gravelly lithofacies

concentrated in thin horizons. Ostracods are present even if scanty. Th e uppermost part of the deposit is composed of blackish clays lacking molluscs, rich in organic matter and sulphur concretions. Th is lithofacies is sharply overlain by Zanclean marine clays of sub-synthem BS3 (see below).

Th e UM2 unit has been observed exclusively in the Piaggiole A quarry (Fig. 3), coinciding with the Bos-sio et al. (1993) “unit A”. Successively this unit was ascribed to the “Argille del Casino” (Bossio et al. 2001), considered to record a second lacustrine episode (“upper lacustrine cycle”, Lazzarotto & Sandrelli 1977) in the adjacent Casino Basin. Renewed lacustrine condition established in the latter basin following the deformation and erosional truncation of UM1 deposits. Angular unconformity within the lacustrine succession of the Casino Basin represents an “intramessinian” uncon-formity of regional signifi cance (see below).

New survey of the Casino Basin’s upper Miocene succession (Benvenuti pers. comm.), indicated that the UM2 deposits in the Strolla Creek are at least partially separated from such lithologically equiva-lent deposits in the Casino Basin, by a substratum high. Th us the UM2 lithofacies is considered to

be accumulated in the southeastern end of the Valdelsa Basin simultaneously to similar deposition in the second Casino’s lake. Deposits lithologically and paleontologically equivalent to UM2, in fact, have been drilled in the central part (San Miniato area) of this basin (Ghelardoni et al. 1968) sug-gesting that a Lago-Mare-like setting occurred on a large portion of the Valdelsa Basin during the latest Miocene. We consider, therefore that the transition bewteen UM1 and UM2, though not visible, is marked by an unconformity related to the tectonically-driven transition from the early Casino Basin to the early Valdelsa-late Casino Basin system (see Discussion).

BORRO STROLLASYNTHEM (UPPERMOST

MESSINIAN-LOWERMOST ZANCLEAN)

Th e deposits included in this study in the Borro Strolla synthem (BS) were referred by Bossio et al. (1993) to the upper Miocene continental gravels (“Conglomerati di Lilliano”in Bossio et al. 2001) and to “unit B” (“Argille Azzurre-lower part” in Bossio et al. 2001). Th ese deposits are overlain by uppermost Zanclean-Piacenzian Certaldo synthem in

(9)

0 m 0 m 0 m 0 m 0 m 5 Log 1 BS1 BS1 BS1 BS2 BS2 BS2 BS3 BS3 BS3 BS3 BS3 Log 2 Log 3 Log 5 Log 4 5 5 5 5 15 10 10 BS1 BS2 BS3 clay granules sand silt boulder pebble clay granules sand silt boulder pebble Log 1

Log 2 Log 3 Log 4 Log 5

BS1 BS3 BS3 BS1 BS1 BS1 BS2 BS2 BS2 BS1 Bs1 BS1 BS1 BS1 BS3 BS3 BS3 BS3 BS3 BS3 BS3 BS3 BS2 BS2 BS2 BS3 BS3 BS2 approx. 50 m

colluvial deposits cross-stratification

terrestrial molluscs horizontal stratification

microvertebratesmacrovertebrates vegetal remainsnormal faults Borro Strolla synthem

N S clay granules sand silt boulder pebble clay granules sand silt boulder pebble clay granules sand silt boulder pebble

FIG. 5. — Photographic sketch, line drawing and correlation of logged sections in the lower-mid portion of the Borro Strolla synthem exposed in the Piaggiole B quarry face (see text).

turn capped by the Piacenzian Ponte a Elsa synthem which are not discussed in this paper. BS unconform-ably overlies UM1 deposits whereas the stratigraphic contact with UM2 has not been observed. Neverthe-less, we favour an unconformable transition as well, basing on the following considerations:

– UM2 and the Borro Strolla synthem crop out at very short distance respectively in the Piaggiole A and B quarries (Fig. 3). Due to the signifi cant thickness of the gravelly-sandy portion of BS, well-exposed on the left bank of the Strolla Creek (Fig. 4), a stratigraphic onlap of UM2 on these deposits seems highly improbable being, in fact, never documented (see also Bossio et al. 1993, 2001). Furthermore a lateral transition from alluvial gravels and sand into lacustrine muds in a so short distance, appears sedimentologically inconsistent;

– despite BS is aff ected by normal faults (Fig. 3), off set doesn’t prevent the recognition of its internal architecture proving limited tectonic displacement. Nevertheless, a normal fault, visible in the Piaggiole A quarry area, downthrows the BS deposits indicating an original position over the UM2 deposits; – lowermost Zanclean marine clays (Bossio et al. 1993, 2001) onlapping UM2 at the Piaggiole A quarry and the fl uvial, gravelly-sandy portion of the BS elsewhere (Fig. 3 see geologic section), represent in fact, the uppermost portion of this synthem.

We suggest therefore that UM1-UM2 and BS de-posits are separated by an high-relief erosional contact recording a NW-SE trending deeply incised fl uvial valley (see discussion). Th e recognition of low-rank erosional surfaces within the BS deposits allowed to establish three sub-synthems (Figs 3; 5).

(10)

TABLE. 1. — Checklist of micro- and macromammal taxa found in the BS1 and subordinately BS2 sub-synthems.

Artiodactyla Bovidae: Protoryx-Paleorix group Cervidae: cf. Paracervulus Giraffi dae: gen. et sp. indet. Insectivora Erinaceidae: Galerix cf. depereti Rodentia Cricetidae: Apocricetus barrierei

Gerbillidae: Debruijnimys sp. Muridae: Apodemus gudrunae

Centralomys benericettii Stephanomys aff. donnezani

Sciuridae: gen. et sp. indet. Lagomorpha Ochotonidae: Prolagus sorbinii

Leporidae: gen. et sp. indet.

Sub-synthem BS1

It is composed of sands and subordinate gravels observed exclusively in the Strolla Creek valley at the base of the Piaggiole B quarry face (Fig. 5). Th e lithofacies association, 1.5-9 m thick, is represented by an alternation of greyish-purple coarse-grained sands, silty sands, and subordinate gravels. Sands range from normal graded, horizontal planar and high-angle cross laminated, to massive in places. Th e high-angle cross lamination indicates palaeocurrent from S-SE. Gravels are polymodal, clast-supported with abundant sandy-silty interstitial matrix in dm-m thick lenticular beds interbedded with sands. Clast composition is mostly represented by lime-stones deriving from the Ligurid unit that occurs on the northeastern margin, and subordinately by sandstones, limestones and scanty quartzites sup-plied from the Tuscan unit (Macigno Fm., Calcare

Cavernoso, Verrucano). Sands and silty sands bear

non-marine mollusc and vertebrate fossil fauna (see below). Th is unit is unconformably overlain by the sub-synthem BS2 deposits.

Sub-synthem BS2

Th e lithofacies association included in this sub-synthem is exposed exclusively in the Piaggiole B quarry (Fig. 5) and consists of gravels grading up-ward to coarse-grained sands, silty sands and silty clays. Gravels are polymodal, clast-supported with abundant grey sandy matrix and showing a clast composition similar to sub-synthem BS1. Clast im-brication indicates a provenance from SE. Th e sands

are massive fi ning upward to silty sands and dark grey silty clays. Th e latter are characterised by dispersed organic matter, root traces, land and subordinate freshwater molluscs and scanty vertebrate remains (see below). Th ese deposits are unconformably overlain by sub-synthem BS3 deposits.

Sub-synthem BS3

Th is sub-synthem consists of two vertically stacked lithofacies assemblages well-exposed on the Piag-giole B and Sefi quarry faces (Figs 4; 5):

– the lower lithofacies assemblage is mostly represented by gravely and subordinate sandy deposits arranged in two main fi ning-upward bedsets separated by an erosive surface. Gravels are polymodal, well to moderately, clast-supported with abundant coarse grained sandy matrix. Locally gravels are characterised by trough and planar cross-bedding indicating paleocurrent to W-SW. Th e sands are coarse-medium grained showing horizontal and trough cross-stratifi cation;

– the upper lithofacies assemblage rests on the previous one through a sharp surface marked by oxidized iron encrustations, gypsum crystals and concretions. Furthermore it sharply overlies the UM2 just at the Piaggiole A quarry (see also Bossio

et al. 1993, 2001). Th e lithofacies assemblage, up

to 40 m thick, is represented by massive grey silty clays and silts becoming dominant toward the top. Th ese deposits are characterised by dispersed organic matter and vegetal debris and by a rich fossil marine invertebrate fauna including molluscs, ostracods (see below) forams and nannofl oras (Bossio et al. 1993, 2001). Bossio et al. (1993, 2001) reported the results of a biostratigraphic analysis of this deposit (“unit B” = Argille Azzurre-lower part) suggesting an earliest Pliocene age (Discoaster variabilis-

Spaeroi-dinellopsis seminulina s.l. biozones).

PALEONTOLOGICAL ANALYSES

Th e material is house in the Science Department of the University of Firenze.

MAMMALS

A not abundant mammal assemblage has been recovered at Strolla Creek, with at least 12 taxa

(11)

FIG. 6. — Small mammals recovered from Borro Strolla synthem: A, Galerix cf. depereti Crochet, 1986, left M3 (DST-BS-07); B,

Apocri-cetus barrierei Mein & Michaux, 1970, left M1 (DST-BS-05); C, Debruijnimys sp. right mandible with M1-M2 (DST-BS-08); D, Apodemus cf. gudrunae van de Weerd, 1976, right M2 (DST-BS-09); E, F, Centralomys benericettii De Giuli, 1989, left M1 and M2 (DST-BS-10); G, Stephanomys aff. donnezani Depéret, 1890, left M1 (DST-BS-11); H, Prolagus sorbinii Masini, 1989, right P3 (DST-BS-06); I, Sciu-ridae indet., M3 (DST-BS-12). Scale bar: 1 mm.

A B D G E F H I C

documented (Table 1). Fossils, on the whole col-lected in the Piaggiole B quarry, come mostly from deposits of the BS1 sub-synthem (from sediments of the BS2 sub-synthem, only remains of the ocho-tonid Prolagus Pomel, 1853 are occurring).

Th e greater part of fossils belongs to micro mammals (Fig. 6) with an insectivore, six rodents and two lagomorphs; the ochotonid Prolagus sorbinii Masini, 1989 (Fig. 6H) is the best represented taxon.

Th e completeness of the mammal fossil record is low and both the large and small mammal samples cannot be considered fully representative of the original assemblage. Th e contribution of mammals to the palaeoclimatic and palaeoenvironmental recon structions is therefore of limited value, taking also into account that some taxa are not completely located in a precise systematic scheme; on the other hand more signifi cant are the biochronological setting and the paleobiogeographic considerations deriving from the mammal remains.

Th e Strolla Creek mammal assemblage, suggests palaeobiogeographic connections with western Europe, fi tting the general aspect of a fauna at the Mio-Pliocene boundary. Indeed, the cervid

Para-cervulus Teilhard de Chardin & Trassaert, 1937

(Fig. 7A), one of the younger representatives of the Muntiacinae group in Europe and one of the better known cervids in France during the Ruscin-ian Mammal Age (Dong 1990, 1996; cf. Azanza 2000), occurs in some Italian localities referred to the latest Turolian (e.g., Baccinello V-3, Abbazzi 2001; Monticino quarry and Casino Basin, Abbazzi & Azanza 2000; Gallai 2002). Th e bovids Palaeoryx Gaudry, 1861 and Protoryx Forsyth Major, 1891 (Fig. 7B, C) well-known from Greek and Turk-ish localities during the middle and late Miocene (Solounias 1981; Köhler 1987), are also reported in Spain (Alcalà 1994) during the latest Miocene and early Pliocene.

Th e occurrence of a very large giraff e (Fig. 7D, E), comparable in size to Helladotherium Gaudry, 1860 from Greece, is an interesting new datum because this group of ruminants is not known from not en-demic Neogene Italian sites. However, it does not allow any biochrological conclusion. In fact, large sized giraff es, widespread during the late Mio cene, are still present in the Pleistocene in eastern Europe (cf. Geraads 1998a), while the occur rence of a dubi-ous sivatheriine is reported from France in the early Pliocene (Montpellier, MN14 unit; Guerin 1986).

(12)

Also the micromammal fauna shows a transitional character (Rinaldi 2003) partially comparable with other latest Messinian mammal assemblages such as those from Monticino, in Central Italy (Marabini & Vai 1989). Indeed some of the taxa occurring in Strolla Creek are also present in the fi llings of the karst fi ssures in the gypsum within the Monticino quarry (De Giuli 1989; Masini 1989; Rook 1992), correlated to the latest Messinian Colombacci Fm: Galerix aff . depereti Crochet, 1986 (Fig. 6A),

Apocricetus cf. barrierei Mein & Michaux, 1970

(Fig. 6B), Apodemus cf. gudrunae van de Weerd, 1976 (Fig. 6D), Centralomys benericettii De Giuli, 1989 (Fig. 6E, F) and Prolagus sorbinii Masini, 1989 (Fig. 6H). However, the murid Stephanomys Schaub, 1938, which is documented at Monticino by the species S. debruijni De Giuli, 1989, is represented at Strolla Creek by remains (Fig. 6G) which are closer to the Pliocene species S. donnezani Depéret, 1890 (e.g., from the French site Sète; Adrover 1986), and another important diff erence is the occurrence of a representative of the Rodentia family Gerbillidae Gray, 1825 (Fig. 6C); the latter form needs a more extensive discussion. Th e dispersal of these rodents from Africa towards Spain is indeed one of the main events related to the Messinian Salinity Crisis (Wessel 1998, 1999; Agustí 1999).

Th e morphology of the gerbil remains from Strolla Creek is comparable to that of the genus

Debruijni-mys Castillo & Agustí, 1996. Th is genus has been

defi ned by Castillo & Agustí (1996) on the material from the middle Pliocene locality Asta Regia (MN 15 biochronological unit), which has been referred to the species D. julii Castillo & Agustí, 1996.

De-bruijnimys occurs also as an indetermined species

(Debruijnimys sp.) in other latest Miocene-early Pliocene Spanish localities (Bacochas 1, cf. Sesé 1989; Caravaca 1, cf. Bruijn 1974; Alcoy, cf. Castillo & Agustí 1996; Agustí 1999; Agustì & Casanovas-Vilar 2003). An origin of the European species of

Debru-ijnimys by migration from North Africa during the

late Messinian is further supported by the occurrence of D. davidi (Geraads, 1998) in the latest Miocene/ early Pliocene locality Lissasfa (Morocco) (Geraads 1998b; Agustí & Casanovas-Vilar 2003).

Th e Strolla Creek gerbil shows a mosaic of char-acters which does not occur in the African sample,

nor in the Spanish ones. As a matter of fact, the large size – the length of M1 falls at the largest extreme of the range of the type sample of D. julii from Asta Regia and it is signifi cantly larger than the African species – and the large anteroconid are associated to the well-developed posterolofi d and to the very brachyodont dental crown (Fig. 6C). Taking into account the unknown morphological variability and the possibility of endemism, we refer the Strolla Creek gerbil to Debruijnimys sp.

Th e occurrence of gerbils in the Neogene of Italy is not, however, a novelty: Gerbillidae gen. et sp. indet. is fi rstly reported from the latest Messinian Ciabot Cagna site (Piedmont, Cavallo et al. 1993). However, the scarcity of the material – only a M3 has been been found at Ciabót Cagna – prevents any comparison with our remains.

To sum up, we believe that the occurrence of gerbil attributable to the genus Debruijnimys, as well as its association with the murid Stephanomys, represented by a species (Stephanomys aff . donnezani) more evolved than the latest Messinian one, could be indicative of a latest Messinian-earliest Zanclean age for the BS1 sub-syntem.

Very general palaeoenvironmental inferences can be drawn from the mammalian assemblage, which are however in agreement with those derived from other sources of informations (e.g., molluscs, see below). Indeed an environment with wooded areas and more open spaces is suggested by the occur-rence of the cervid on one hand, and of the large sized giraffi d and Debruijnimys sp. on the other; modern gerbils are in fact widespread in the desertic and sub-desertic regions of Africa and central and western Asia (cf. Tong 1989).

MOLLUSCS

UM2 deposits

A signifi cant rich oligo-mesohaline mollusc assem-blage is recorded from unit UM2 sampled in the Piaggiole A quarry (Fig. 8). It is composed of aquatic prosobranch gastropods belonging to the families Neritidae Rafi nesque, 1815, Melanopsidae H. & A. Adams, 1854 and Hydrobiidae Troschel, 1857 and bivalves belonging to the families Cardiidae Lamarck, 1809 (Lymnocardiinae Stoliczka, 1871) and Dreissenidae Gray in Turton, 1840.

(13)

A

B

D

E

C

FIG. 7. — Large mammals recovered from Borro Strolla synthem: A, Paracervulus sp., left M1, occlusal view (DST-BS-02); B, C, Paleoryx group (DST-BS-03); B, frontals with horn cores; C, left metatarsus; D, E, Giraffi dae indet., distal end of metapod (DST-BS-04). Scale bars: A, D, E, 2 cm;B, C, 10 cm.

(14)

Th e gastropods are represented by few species with a high number of specimens: Th eodoxus mutinensis

D’Ancona, 1869, Saccoia etrusca Capellini, 1880,

Saccoia fontannesi Capellini, 1880 and Melanopsis

narzolina D’Archiac, 1846. Th e bivalves are

domi-nated by several species of Lymnocardiinae:

Euxini-cardium subodessae Sinzov, 1877, Pseudocatillus cf. P. pseudocatillus Barbot de Marny, 1869, Pontalmyra partschi Mayer, 1871, Chartoconcha sp., Prosodac-nomya sturi sabbae Andreescu, 1975, Psilodon cf. P. haueri Cobalcescu, 1883, Psilodon cf. P. munieri

Stefanescu, 1896 and other taxa in a fragmentary state of preservation. Dreissenidae are represented by one species: Dreissena ex gr. D. rostriformis Des-hayes, 1832.

Th e gastropods T. mutinensis, S. etrusca and

S. fontan nesi are typical extinct endemic elements,

occurring in hypo-oligohaline oxygenated shallow water enviroments of Italian basins from the late Tortonian-early Messinian to the latest Messin-ian (Ghetti et al. 2002 and references therein).

Melanopsis narzolina is widespread in the

hypo-oligohaline enviroments of several upper Miocene basins of the Mediterranean area and in Italy where it occurs in latest Messinian deposits; it seems to be also present in lower Pliocene deposits of east-ern Europe (Romania and Slavonia) (Wenz 1929, 1942; Esu 1980).

Th e subfamily Lymnocardiinae is very important for palaeobiogeographic inference, being its numer-ous representatives widespread since the Oligocene in the western and eastern Paratethys (Nevesskaya

et al. 2001). Th e genera and species belonging to

this subfamilty recorded from Strolla Creek deposits are of particular palaeoecologic, chronostratigraphic and palaeobiogeographic interest. Th ey charac terize the latest Messinian Lago-Mare biofacies with low salinity (oligo-mesohaline) and shallow water oc-curring in several latest Messinian deposits (such as Colombacci Fm.) of the Italian peninsula and Sicily during the post-evaporitic phase (Esu 2002, 2007). Moreover close relations with the lower “Pontian” (Odessian) fauna of the Aegean area and strong affi nity with Paratethyan assemblages of Pontian age from Dacian and/or Euxinian ba-sin (Andreescu 1977; Popov & Nevesskaya 2000; Nevesskaya et al. 2001) are pointed out by several

recorded Italian Lymnocardiinae taxa (Di Geronimo

et al. 1989; Esu 2007). Dreissena ex gr. D. rostri-formis, an upper Miocene species of Paratethyan

origin (Archambault-Guezou 1976), also points to oligohaline oxygenated waterbody. Th e Paratethyan genera Chartoconcha Andrussov, 1908 and Psilodon Cobalcescu, 1883 are new for the Italian fauna. Since the ecology of Lymnocardiinae and Dreisse-nidae is linked to low salinity, spreading of suitable habitat in depositional systems of marginal settings characterized by increasing freshwater infl ux after the “salinity crisis” favoured their dispersal into the Mediterranean area from the Paratethyan realm. Indeed the well-preserved and diversifi ed mollusc assemblage from the Strolla Creek UM2 deposits indicates the occurrence of a shallow water oligo-mesohaline environment characterized by a latest Messinian “Lago-Mare” biofacies in that area.

Sub-synthems BS1 and BS2

Sands of sub-synthem BS1 and the uppermost muds of BS2, both sampled in the Piaggiole B quarry, bear non-marine mollusc assemblages character-ized by some pulmonate gastropods. Th e common recorded species are Carychium tridentatum Risso, 1826, Physa sp., Acanthinula aculeata O. F. Müller, 1774, Cernuella sp., whilst Parmacella sp.,

Trunca-tellina callicratis Scacchi, 1833, a fragment of the

aquatic prosobranch Melanopsis sp., is exclusive of BS1, and Oxychilus sp. and Cepaea sp., have been found only in BS2.

On the whole this fauna is composed of species of diff erent habitats with a medium degree of aridity and developed under temperate-warm conditions:

C. tridentatum and A. aculeata occur commonly in

damp and wooded places, whilst T. callicratis,

Parma-cella sp. and Cernuella sp. are elements living in open

drier places (Kerney & Cameron 1979). Physa sp. and Melanopsis sp. require slow running-water. Such type of assemblage, in which the terrestrial taxa are dominant, points to mollusc reworking in an envi-ronment liable to fl ooding. Th e recorded species are known so far since the (early?) Pliocene to Recent (Esu & Ciangherotti 2004) except for the genus

Parmacella Cuvier, 1804 which is known in Italy

from Messinian to Early-Middle Pleistocene and is still living in warmer Mediterranean and Middle

(15)

A

E

B

D

C

F

G

H

K

J

I

L

FIG. 8. — Non-marine molluscs from UM2 deposit: A, Euxinicardium subodessae Sinzov, 1877, left valve; B, C, Pseudocatillus

pseudocatillus Barbot de Marny, 1869, right valve, interior and exterior view; D, Pontalmyra partschi Mayer, 1871, right valve; E, F, Prosodacnomya sturi sabbae Andreescu, 1975, right valve, exterior and interior view; G, Chartoconcha sp., right valve; H,

Psilo-don cf. P. haueri Cobalcescu, 1883, right valve; I, Psilodon cf. P. munieri Stefanescu, 1896, left valve; J, Lymnocardiinae indet., right valve. K, L, non-marine molluscs from BS2 deposit; K, Oxychilus sp., apical view; L, Cepaea sp., apical view. Scale bars: A-C, 3 mm; D-F, J, 4 mm; G-I, 15 mm; K, 1 mm; L, 7 mm.

(16)

East areas (Manganelli & Giusti 1990), as well as the genus Melanopsis Férussac, 1807, pointing to warm climate and a semi-arid environment. A prob-able trend from semi-arid and warm conditions to a more humid climate can be hypothesized for the time-span deposition of BS1 and BS2, taking into account the occurrence of the genera Parmacella,

Truncatellina Lowe, 1852 and Melanopsis only in

the fi rst sub-synthem BS1.

OSTRACODS

Upper Miocene deposits

Th ree samples have been collected in very small outcrops referable to late Miocene, respectively located NE and SE of the Strolla Creek: two came from unit UM1 (samples 1 and 2, Fig. 3), whereas the third has been collected at Piaggiole A quarry within unit UM2.

UM1 deposits. UM1 samples bear well preserved

and rather abundant ostracods (Fig. 9A-E), each species represented both by adults and instars, thus considered in situ species. Th e ostracod assemblage from sample 1 is made of brackish Cytheroidea Baird, 1908 such as Cyprideis sp. 1, Cyprideis sp. 2,

Cyprideis sp. 6 and Loxoconchissa (Loxocaspia) sp. 1.

Th e ostracod assemblage from sample 2 consists, on the contrary, mainly of Cypridoidea: Candoninae Kaufmann, 1900 (Lineocypris spp., Candona

[Negle-candona] sp., Fabaeformiscandona sp., Labiatocan-dona sp. and PseudocanLabiatocan-dona sp.), Cyclocypridinae

Kaufmann, 1900, (Cypria sp. and Physiocypria sp.) while Cytheroidea are limited to Cyprideis sp. 6 and Tavanicythere joachinoi Bossio, Gliozzi & Tas-sone, 2004.

Despite the open nomenclature of some spe-cies, the recognized assemblages suggest the ex-istence of a brackish waterbody characterized by slight fl uctuation in salinity. Cyprideis Jones, 1857, Tavanicythere Bossio, 1980 and Loxoconchissa

(Loxo-caspia) Triebel & Malz, 1969, represent typical

brackish water dwellers, whereas candonids such as

Lineocypris Zalanyi, 1929, and Labiatocandona sp.

(both typical of upper Miocene Paratethyan depos-its) tolerate slightly saline waters. Th ese ostracod assemblages are referable to the late Tortonian-early Messinian (Bossio et al. 2001) and confi rm the

correlation of UM1 with the deposits of the fi rst lacustrine episode in the Casino Basin (Lazzarotto & Sandrelli 1977). Th ey document that the fi rst lake of the Casino Basin signifi cantly extended to the NW. Th e UM1 ostracod assemblages are dominated by western European and Mediterranean species in full agreement with the late Tortonian-early Mes-sianian ostracods recorded in other Tuscan basins. Th e occurrence of Lineocypris spp. and

Labiato-candona sp., pertaining to the Paratethyan domain

and not exclusive of UM1 (Belforte and Velona basins, Devoto 1968; Ghetti et al. 2002; Volterra-Radicondoli Basin, Bossio et al. 1981; Krstic & Bossio 1992; Testa 1995; Gliozzi et al. 2007), is tentatively explained by the passive dispersal via aquatic birds from the Paratethyan regions (Ben-son 1976; Ben(Ben-son & Rakic-El Bied 1991; Gliozzi

et al. 2007).

UM2 deposits. Th e Piaggiole A quarry ostracod as-semblage (sample 3, Fig. 3) is made of few valves of

Caspiocypris cf. C. pontica Sokac, 1852, Cyprideis sp., Loxocorniculina djafarovi (Schneider in Suzin, 1906)

and Loxoconcha sp. (Fig. 9F-J).

Th e species detected in this sample again point to a brackish (oligo- to mesohaline) waterbody, but the presence of two typical Pontian/Dacian Paratethyan species such as Caspiocypris cf. C.

pon-tica and L. djafarovi indicates a very restrict time

interval corresponding to the latest Messinian Lago-Mare biofacies, widespread almost all over the Mediterranean Basin (Carbonnel 1978; Cipollari

et al. 1999a, b; Gliozzi 1999; Gliozzi et al. 2002,

in press). In particular, the presence of L. djafarovi seems to constrain the age of the deposits to the last latest Messinian precessional cycle (Gliozzi et al. 2006). Th ese results, in full agreement with those of Bossio et al. (1993), which reported a richer Para-tethyan ostracod association from these deposits, suggest a palaeoenvironmental and chronological correlation with the Argille del Casino and confi rm the develop ment of the Lago-Mare realm east of the Mid Tuscan Ridge. Diff erently from the basins located to the west, where the Lago-Mare followed to pre-existing restricted marine environments (Volterra- Radicondoli Basin, Bossio et al. 1978; Bossio et al. 1981, 1996a, b; Sarti & Testa 1994;

(17)

A B C F G H I J D E

FIG. 9. — Ostracod valves from UM1 (samples 1, 2, Fig. 3A), UM2 and Borro Strolla synthem-upper part (samples Piaggiole A quarry, Fig. 3B): A-C, sample 1; A, Cyprideis sp. 1, male left valve in lateral external view; B, Cyprideis sp. 2, female right valve in lateral external view; C, Cyprideis sp. 6, male right valve in lateral external view; D-F, sample 2; D, Labiatocandona sp., carapace in left lateral view; E, Tavanicythere joachinoi Bossio, Gliozzi & Tassone, 2004, female right valve in lateral external view; F-J, sample Piaggiole A quarry; F, Loxocorniculina djafarovi Schneider in Suzin, 1906, female left valve in lateral external view; G, Caspiocypris cf. C. pontica Sokac, 1852, female right valve in lateral external view; H, Henryhowella sarsi Müller, 1894, left valve in lateral external view; I, Bythocypris

obtusata producta Seguenza, 1880, left valve in lateral view, transmitted light; J, Krithe padovani Colalongo & Pasini, 1988, left valve in lateral view, transmitted light. Scale bars: 0.1 mm.

(18)

Fine Basin, Aldinucci et al. 2005), in the eastern basins, fully or partially isolated from the marine infl uence, brackish lakes may have been favoured by: 1) the infl ow of saline water generated by the dissolution of the Triassic evaporites (isotopic studies by Anadón et al. 2002 in the lower Messinian of the Velona Basin); or 2) a partial connection with the Lago-Mare facies, in the specifi c case possibly exist-ing in the western portion of the Valdelsa Basin as suggested by core data (Ghelardoni et al. 1968).

Sub-synthem BS3

No ostracod valves have been recovered in sub-synthem BS1, BS2 and in the lower part of the sub-synthem BS3 deposits. One sample from the mudstone on top of sub-synthem BS3 collected in the Piaggiole A quarry, provided a scanty but well-preserved marine ostracod assemblage, made of Bythocypris obtusata producta (Seguenza, 1880),

Palmoconcha agilis Ruggieri, 1967, Krithe padovani

Colalongo & Pasini, 1988, Henryhowella sarsi Müller, 1894 and Macrocypris Brady, 1868. Th is associa-tion is very similar to that reported by Bossio et al. (1993, 2001) from the same deposits.

Th e lower diversity of this assemblage coupled with the presence of marine taxa such as Krithe Brady, Crosskey & Robertson, 1874, Bythocypris Brady, 1880 and Henryhowella Puri, 1957 suggest a rather deep (outer circalittoral/bathyal) marine environ-ment. Th e co-presence of P. agilis (early Messinian [Ruggieri 1967b; Aruta 1982] to Zanclean [MPl 2 or G. margaritae zone] [Nachite 1993]), Bytho cypris

obtusata producta (middle Mliocene-early Pleistocene

[Aiello et al. 2000]), H. sarsi (middle Miocene-Present [Bonaduce et al. 1999]) and K. pado vani (Zanclean [MPl 2 or G. margaritae zone] [Ciampo 1992] to Santernian [MPl 6 or G. infl ata zone] [Colalongo & Pasini 1988])point to the Early Zanclean. Due to the relatively low resolution of the Pliocene marine ostracod biostratigraphy, this age is considered to be in an overall agreement with that proposed by Bossio et al. (1993, 2001), based on the nannoplankton and planktic foraminifer assemblages.

Finally, the composition of this assemblage, made of a typical Mediterranean-Atlantic stock, attests to the completely restored connection between

Atlantic and Mediterranean at the beginning of the Pliocene.

DISCUSSION

Th e stratigraphic revision of the succession encom-passing the Mio-Pliocene transition in the Strolla Creek area allows to focus the discussion on two specifi c aspects.

LOCALDEPOSITIONALEVOLUTION:

REGIONALSEA-LEVELRISEVSLOCALTECTONISM

Th e stratigraphic relations among the diff erent units recognized in this study illustrate a complex depositional dynamic. Non-marine, fl uvial and marine conditions established in the Strolla Creek area in the following sequence (Fig. 10).

Th e late Tortonian-early Messinian deposition

and the intramessinian tectonic event

Unit UM1, bearing scanty molluscs and meso-/ oligohaline ostracod assemblages, accumulated during the fi rst fl uvio-lacustrine episode of the Casino Basin in the late Tortonian-early Messinian (Lazzarotto & Sandrelli 1977; Bossio et al. 2001; Fig. 10A). Th is evidence indicates that the fi rst Ca-sino expanded to NW fl ooding an area that succes-sively will pertain to the Valdelsa Basin. Th e upper Messinian UM2 deposits record the development of a non-marine fl ooded environment (Fig. 10B) quite similar to those developing in the same time-span in other Tuscan basins. In particular such an environment occurred in the nearby Casino (e.g., during the second lacustrine episode), Siena basins (Lazzarotto & Sandrelli 1977; Bossio et al. 2001) and in the central part of the Valdelsa Basin (Ghelardoni

et al. 1968). Deposition was characterised by

sedi-ment settling in quiet shallow waters populated by a diversifi ed benthic invertebrate fauna. Th e mol-lusc and ostracod assemblages characterising these deposits show a marked Paratethyan affi nity being referable to the latest Messinian Lago-Mare biofacies (Esu 2002 and references therein; Gliozzi et al. 2002 and references therein). Th e attribution of UM2 to the Valdelsa Basin rather than to the Casino Basin points to a physical separation established between

(19)

the two areas after the deposition and deformation of UM1-“lower lacustrine cycle”/“Argille del

Tor-rente Fosci”. Th is “intramessinian” tectonic event

(see below) caused the development of a second lacustrine setting in the Casino Basin and the possible activation of the SE Valdelsa Basin where “Lago-Mare” condition determined the deposition of UM2 and similar deposits drilled in the central portion of this basin. In particular the physical separation of the Casino from the Valdelsa Basin possibly testifi es to an actively growing structural high tranversal to the basin’s axes.

Th e combination of subsidence pulses and rapidly

rising sea-level at the Messinian-Pliocene transition

In the latest Messinian a vigorous fl uvial rejuvena-tion caused the deep incision of the UM1-UM2 deposits with the development of a N-S trending fl uvial valley about 2 km wide (Fig. 10C). Th e western bank of this valley, although disturbed by faults, is located between Piaggiole A and B quar-ries whereas the eastern bank is approximately on the right side of the Carfi ni Creek around Fizzano (Fig. 3). Th e valley is fi lled by the Borro Strolla synthem accumulated throughout three major stages of alluvial deposition. Palaeocurrent data indicate the occurrence of tributaries which during the deposition of BS1 and BS2 supplied sediments from the SE to the main valley oriented NNE-SSW. Th e latter drained toward the SW (Colle Valdelsa area) as indicated by palaeocurrent in BS3.

Th e BS1, BS2 and part of BS3 sub-synthems record three periods of valley aggradation punctuated by erosion (Fig. 5). Th e internal lithofacies architec-ture in the three sub-synthems is characterised by fi ning upward trends related to signifi cant change of sediment supply and base level.

BS1 sub-synthem records a transition from high to low-energy channels fi lled with gravels abruptly grading into sands and silty sands. Th e non-marine molluscs point to slow fl owing water and to an environ ment liable to fl ooding. A similar conclusion can be drawn from the analysis of mammal fossils. Th e low abrasion showed by some bones and, in particular, the presence of articulated limb bones indicate that burial occurred soon after the death of specimens and that fl uvial transport did not play

signifi cant role in the setting of mammal fossils. In BS2 the transition from a gravel-bed river to low-energy environment is outlined by organic matter-rich and terrestrial mollusc-bearing muds. Th ese deposits are interpreted as accumulated in a fl oodplain marking the deactivation of the coarse-grained transport and deposition. In the thicker BS3 sub-synthem two gravel-sand depositional units record as many phases of decreasing bed-load grain size which refl ect a cyclic shift of the fl uvial style, from gravel- to sand-bed river, as a consequence of reduced sediment supply and/or reduced transport capacity, both phenomena com-patible with a relative base-level rise (Shanley & McCabe 1994). In the studied case fi ning-upward trend of successive valley fi lls may refl ect the fl u-vial response to uplift/denudation cycles (Blair & Bilodeau 1988; Paola et al. 1992; Mutti et al. 1996) interfering with a regional sea-level rise (see below). Th e uplift/denudation component determined coarse-grained sediment supply to the fl uvial system during tectonic quiescence whereas sediment starving and fi ning derived from the cou-pled eff ect of active uplift of the Chianti Mounts and consequent subsidence of the Valdelsa Basin. Th e joined eff ect of subsidence pulses and rapidly rising sea-level at the Miocene-Pliocene transition determined periods of high relative base-level in the valley controlling the cyclic, fi ning-upward, fi lling pattern.

A similar, fl uvial, succession occurs in the Casino Basin (Benvenuti et al. pers. comm.) attesting to the possible coeval development of a fl uvial system draining the former Casino Basin and fl owing to the ESE (i.e. toward the Siena Basin). It appears therefore that at the Miocene-Pliocene transition a watershed between the former Casino and the Valdelsa basins controlled a diverging drainage. Such a watershed formed as the consequence of the actively growing tranversal high which conditioned the Lago-Mare sedimentation in adjacent basins during the late Messinian.

Th e fl ooding of the Strolla Creek area at the early

Zanclean maximum sea-level rise

Th e abrupt marine fl ooding marked by the mollusc-bearing mudstone sharply resting on UM1-UM2

(20)

and BS3 fl uvial deposits eff ectively records the maxi-mum Zanclean rise of the sea level that approached the Chianti Mounts (Fig. 10D). An inner shelf, dominated by sediment settling and populated by diversifi ed benthic communities, replaced the fl uvial environment. Th e lack of transitional environments between the fl uvial and the inner shelf and the pres-ence of a circalittoral/bathyal ostracod assemblage indicate a very rapid marine fl ooding.

Th e high sea level recorded by these deposits, reasonably related to the defi nitive re-fi lling of the Mediterranean Basin, may have been the last in a succession of fl ooding pulses which concurred to control the facies stacking pattern of the incised valley fi ll (see Discussion above). Th e physiography of the wide region including central Tuscany and the Northern Tyrrhenian Sea, was characterized in the Late Neogene by several basins separated by structural highs and located at elevations progres-sively higher toward the Northern Apennines divide. Zanclean sea fl ooding of the innermost portion of this region, thus, could have been delayed in respect with the fl ooding of the lowermost Northern Tyr-rhenian Basin and adjoined coastal areas. Such a “threshold eff ect” on sea-level rise, hence on fl uvial base level, may have represented a further control-ling factor for the cyclic sedimentary fi lcontrol-ling of the Borro Strolla palaeovalley.

THE MESSINIANANDTHE MIO-PLIOCENEEVENTS:

FROMALOCALTOAREGIONALPERSPECTIVE

Th e local stratigraphic architecture of the succes-sion recording the Mio-Pliocene transition provides arguments for a comparison of events at wider scales. In general terms the Strolla Creek succession records a discontinuous transition from lacustrine to fl uvial to open marine settings evidenced by two signifi cant internal unconformities. Th e basal one separates UM1 from UM2 deposits whereas the upper unconformity subdivides UM1-UM2 from the Borro Strolla synthem.

Occurrence of unconformities in the Messinian successions of the Tuscan basins, documented on both sides of the Mid Tuscan Ridge, is discussed for comparison with the studied case (Fig. 11).

At least three major “intramessinian” unconformi-ties are recognized in the Radicondoli-Volterra Basin

(Fig. 1A) located west of the Mid Tuscan Ridge and bearing evaporites. Here, the pre-evaporitic terres-trial to shallow marine deposits, including primary evaporites pre-dating the Salinity Crisis (Benvenuti

et al. 1999; Testa & Lugli 2000), are deformed (see

Moratti & Bonini 1999 and Brogi & Liotta 2005 for alternative structural interpretations) and un-conformably overlain, through unconformity Ima, by clastics and resedimented evaporites. Angular unconformity Imb (Testa & Lugli 2000; Brogi et

al. 2002, see also Pascucci et al. 2004 for a diff

er-ent inter pretation of this stratigraphic transition) separates two gypsum units which are referred respectively to the lower (primary) and upper (re-sedimented) evaporites of Sicily and Spain (Testa & Lugli 2000). Finally unconformity Imc puts in

con-tact the upper gypsum unit with a stratigraphically complex clastic succession including resedimented evaporites, gravels, sands, marls and clays (Monte-mamboli Group in Testa & Lugli 2000). On the whole the upper gypsum and overlying deposits are referred to the Lago-Mare realm (Testa & Lugli 2000; Pascucci et al. 2004).

East of the Mid Tuscan Ridge, where the Messin-ian is represented by fully terrestrial, evaporite-free, terrigenous successions, a prominent intramessinian unconformity is also documented. In the Velona Basin (Rook & Ghetti 1997; Bonini et al. 1999; Fig. 1B), strongly deformed fl uvio-lacustrine de-posits ascribed to the early Messinian (Ghetti et

al. 2002) are angularly overlain by a further

al-luvial succession possibly of late Messinian age. As mentioned above a similar unconformity has been described in the Casino Basin (Lazzarotto & Sandrelli 1977; Bossio et al. 2001) separating the deformed upper Tortonian-lower Messinian fl uvio-lacustrine deposits from the second fl uvio-uvio-lacustrine succession encompassing the late Messinian. Due to the strong deformation of the underlying upper Tortonian-lower Messinian units, unconformities in these basins are tentatively correlated with the Ima of Volterra-Radicondoli Basin, thus pre-dating the Salinity Crisis. Similarly, we equate the inferred unconformable contact between UM1 and UM2 in the Strolla Creek area to unconformity Ima.

Th e unconformity separating UM1-UM2 from the Borro Strolla synthem documents a further

(21)

? ? ? ? ?? ? ? ? ? ? ? ? ? ? early Zanclean A B C D N N N N Casino first lake

Quer cegr

ossa V alley

late Tortonian-early Messinian

late Messinian transversal gr

owing high

latest Messinian-earliest Zanclean Strolla Cr eek V alley Quer cegr ossa V alley Elsa sea Elsa sea Siena sea ElsaLago-Mare CasinoLago-Mare (secondlake) MONTAGNOLASENESE CHIANTIMOUNTS ElsaArbia-SerbiaBasi ns waterhed ElsaArbia-SerbiaBasi ns waterhed

FIG.10. — A, The Strolla Creek area ( ) is included in the lacustrine Casino Basin; B, a NE-SW trending structural high separates the Casino from the Valdelsa basins, black arrows indicate major supply points; C, the NE-SW high is a watershed for drainage to the Valdelsa and Siena basins respectively; D, at the early Zanclean maximum of sea-level rise the Strolla Creek area is fl ooded.

(22)

event that characterized the end of the Messinian and that can be searched for in other Tuscan basins. Th e stratigraphically complex Montebamboli Group (sensu Testa & Lugli 2000), laying on top of several Messinian successions, has been frequently reported to be conformably overlain by lowermost Pliocene marine mudstones (Bossio et al. 1993). Th e basal unconformity of this succession may be related to the Imc of the Volterra-Radicondoli Basin and, due to the occurrence of several coarse-grained alluvial units possibly separated by low-rank unconformities. On the northwestern margin of the Volterra Basin, for istance, alluvial gravels (i.e. “Conglomerati delle Volpaie”, Lazzarotto et al. 2002), rest unconform-ably on the “Lago-Mare” deposits being in turn overlain by lowermost Pliocene marine mudstones. Th ese alluvial deposits are ascribed to the lowermost Pliocene (Lazzarotto et al. 2002), though a latest Messinian age could be considered.

Similarly, in the Velona Basin Montebamboli-like alluvial deposits, resting unconformably on top of the Messinian succession (Rook & Ghetti 1997), may be reasonably referred to the Mio-Pliocene transition.

As already mentioned, a possibly equivalent un-conformity occurs in the Casino Basin separating the Messinian fl uvio-lacustrine deposits from fl uvial gravels and sands possibly encompassing the Mio-Pliocene transition (Lazzarotto & Sandrelli 1977).

In conclusion we suggest that, despite lacking a precise chronostratigraphy, further low-rank uncon-formities punctuated the late Messinian in response to a generalised uplift of the basin’s margins and that the youngest one, Imd, could have marked

a palaeogeographic transition from lacustrine to alluvial settings in some parts of central-southern Tuscany during the latest Messinian.

Two intramessinian unconformities (Im and Im1), occur also in the evaporite-bearing deposits fi lling the Adriatic foredeep (Roveri et al. 2001, 2003; Fig. 11), and suggest a possible wider signifi cance of tectonic events aff ecting the Northern Apennines during the latest Miocene. Specifi cally, in the fore-deep basin, Im separates primary evaporites from post-evaporitic deposits similarly to the Imb of the

Tuscan basins. Unconformity Im1 subdividing the

post-evaporitic succession in two main units (P-ev1 and P-ev2), can be tentatively equated to Imc-Imd

group of Tuscany. Major tectonic pulses in the North-ern Apennines thrust belt and related deformation of the foredeep are suggested for the origin of these unconformities. Unconformity Im1, in particular, marked the beginning of cyclic sedimentation re-corded in the P-ev2 by fl uvio-deltaic sandstone and mudstones. Th e cyclothemic arrangement, expressed by fi ning-upward trends within an overall backstep-ping fl uvio-deltaic system, is interpreted as the eff ect of precession-dominated climatic cycles which con-trolled water and sediment discharge in the fl uvial systems of the Northern Apennines during the latest Messinian (Roveri et al. 2001, 2003).

A latest Messinian unconformity, is recognized also in other Mediterranean marginal basins (Mo-rocco, Spain, Sardinia, Sicily, Israel; Fortuin et al. 1995; Buchbinder & Zilberman 1997; Rouchy et

al. 2003; Cornée et al. 2004; Soria et al. 2005)

as-sociated to non-marine or fully terrestrial deposits resting on the Lago-Mare deposits. Th is evidence suggests that at a regional scale the Mio-Pliocene transition may have been marked not only by rapid massive sea fl ooding as widely documented in relatively deep basins of eastern Mediterranean and Adriatic regions (Rouchy et al. 2001; Roveri

et al. 2001, 2003) but also by base-level lowering

and consequent fl uvial incision caused by intense tectonic deformation of the circumediterranean margins and associated shallow basins.

CONCLUSIONS

Th e study of a clastic succession referred to the Late Neogene and exposed in the SE portion of the Val-delsa Basin has been carried out by the integration of physical stratigraphy, vertebrate and invertebrate (ostracods and continental molluscs) palaeontology. In particular attention has been focused on the sedi-ments encompassing the Mio-Pliocene transition and included into the Borro Strolla synthem which represents a composite fi ll of a NE-SW trending fl uvial valley, incised in upper Messinian Lago-Mare deposits, culminating with a marine fl ooding. Th e fl uvial deposits characterizing the lower portion of this synthem bear scanty remains of the gerbil

(23)

Gilbert MPL 1 Non distinct. zone Globorotalia conomiozea G. suterae MIOCENE Messinian 6 7 Age (Ma) Chr on. Polarity T ortonian fluvio-lac reef pr-ev res-ev Lago-mare shelf Lago-mare fluvial ? Lago-mare fluvial Volterra-Radicondoli Strolla (SE Elsa) Casino

first Casino lake second Casino lake brackish Mid Tuscan Ridge Northern Apennines divide shelf P. Z. Imc pr-ev res-ev euxinic shales ? p-ev2 p-ev2 p-ev2

p-ev1 res-ev p-ev1 res-evash layer

euxinic shales Marginal basins Deep basins a Deep basins b Adriatic foredeep hiatus shelf euxinic shales ? ? ? ? ? ? ? ? ? ? ? ? ? Im Im1 hiatus hiatus Planktonic Foraminifers Biozone

Imb

Imd

Ima

FIG. 11. — Tentative correlation of Tortonian-Messinian successions across the Northern Apennines region extending the scheme of Roveri et al. (2001) for the Adriatic margin and foredeep basins. Abbreviations: Chron., chronology; P., Pliocene; Z., Zanclean.

S. donnezani (Depéret, 1890) (both rodents with

Pliocene affi nity) which represent signifi cant nov-elties in regard to the latest Messinian mammal fauna Italian record. Th e co-occurrence of terrestrial molluscs from the same sediments with a Pliocene affi nity, gives further biochronologic constrains supporting the calibration of the fl uvial deposits to the latest Messinian-earliest Zanclean. Finally, the marine mudstone on top of the synthem were calibrated in previous studies to the early Zanclean. Besides the revised chronological calibration, the Borro Strolla synthem is characterized by a cyclic facies architecture which is interpreted in terms of interplay between variable sediment supply, depend-ing on local, tectonically-controlled, relief uplift and denudation, and regional rise of sea-level. Th e latter was evidently driven by the earliest Pliocene refi lling of the Mediterranean Basin following to the Salinity Crisis. Th e region between central Tuscany and the Northern Tyrrhenian Sea was characterized

during the Late Neogene by several basins sepa-rated by structural highs determining a “threshold eff ect” on the earliest Pliocene sea-level rise which may have caused pulsatory variation of the fl uvial base level. Th is eff ect may have concurred to the development of cyclic facies trends in the fl uvial portion of the Borro Strolla synthem.

On the whole the Strolla Creek succession records a discontinuous transition from lacustrine to fl uvial to open marine settings evidenced by two signifi -cant internal unconformities which may have some equivalents in the Northern Apennines region. In particular the occurrence of a latest Messinian un-conformity allows to reconsider the local Messinian-Zanclean transition as not properly continuous, a conclusion having possible implications at a wider scale. Th e Mio-Pliocene transition may have been marked not only by rapid massive sea fl ooding as commonly documented in the Mediterranean deep basins. Tectonic uplift in the marginal areas may

Riferimenti

Documenti correlati

La soluzione attualmente praticata dal Comune di Cagliari, che individua gli immobili da assegnare alle famiglie rom nel suo territorio o nel suo hinterland e ne paga le locazioni,

Sebbene la scarsità di notizie sugli okimono e la mancanza di studi specifici abbiano 

Per eseguire le verifiche strutturali è quindi necessario il progetto simulato delle armature dell'intera aula magna, di una parte degli elementi verticali e di quelli orizzontali

Tables 5 and S6 show the results of the correlation analy- sis between the selected NSD-, SR- and SD-type indices for ecophysiological variables and fluxes and biophysical prop-

More specifically, the structure does not tell us whether (1) as uttered now is true simpliciter, for it doesn’t tell us which moment is the present moment.. On the indexical

Differences in the compound relative abundance of VOCs in head plus pro-thorax samples of females and males collected during late spring were analysed by Mann-Whitney U test..

Depression (HRSD) is the de facto standard in symptom assessment and is used in many depression trials identified for overviews. The degree of concordance of the actual effect

Come per tutte le im- prese, anche quelle sociali mostrano mag- giori investimenti di tempo, pensiero e ri- sorse se sono coinvolte direttamente nella vita